Ligand Field Theory

Ligand Field Theory (LFT) is a comprehensive quantum-mechanical framework that describes the electronic structure, covalent bonding, optical spectra, magnetic properties, and structural stability of transition metal coordination complexes. Developed between the 1930s and 1950s by John Hasbrouck Van Vleck, John Stanley Griffith, and Leslie E. Orgel, Ligand Field Theory syntheses two foundational models: it retains the rigorous group-theoretical symmetry analysis of Hans Bethe's electrostatic Crystal Field Theory (CFT) while replacing the idealized assumption of point-charge ligands with Linear Combination of Atomic Orbitals–Molecular Orbital (LCAO–MO) theory.

Whereas pure Crystal Field Theory treats metal–ligand interactions solely as electrostatic repulsion between metal d electrons and negative point charges—an approximation that fails to explain why neutral carbon monoxide (CO) produces a far larger octahedral splitting energy Δ_o than negatively charged fluoride (F⁻)—Ligand Field Theory demonstrates that Δ_o is controlled by covalent σ-donation, π-donation, and π-backbonding between the metal valence (n−1)d, ns, and np orbitals and symmetry-adapted linear combinations (SALCs) of ligand orbitals. Furthermore, LFT reveals that the upper e_g* orbital doublet in an octahedral complex is not merely electrostatically repelled, but is a strongly σ-antibonding molecular orbital—providing the physical foundation for the Jahn–Teller effect, spin crossover thermodynamics, and the nephelauxetic effect in Tanabe–Sugano diagrams.

Interactive VSEPR Lab: Compare octahedral (O_h), square-planar (D_4h), and tetrahedral (T_d) coordination polyhedra in the VSEPR Lab!

Symmetry-Adapted Linear Combinations (SALCs) of Six σ-Donor Orbitals

Consider a regular octahedral complex ML₆ belonging to the O_h molecular point group, where a first-row transition metal ion supplies nine valence atomic orbitals: five 3d orbitals, one 4s orbital, and three 4p orbitals. In O_h symmetry, these nine metal valence orbitals transform as the following irreducible representations:

• 4s: a_1g (totally symmetric sphere)
• 4p (p_x, p_y, p_z): t_1u (triply degenerate, ungerade)
• 3d axial/equatorial (d_z², d_x²−y²): e_g (doubly degenerate, pointing directly along the six M–L axes)
• 3d inter-axial (d_xy, d_xz, d_yz): t_2g (triply degenerate, pointing between the Cartesian axes)

Now consider six σ-donor lone-pair orbitals (σ₁, σ₂, ..., σ₆) located on the six ligands along ±x, ±y, and ±z. Applying the projection operator of the O_h character table decomposes the 6-dimensional reducible representation Γ_σ into three symmetry-adapted linear combinations (SALCs):

Γ_σ = a_1g + e_g + t_1u

Specifically, the six ligand σ orbitals combine with matching phase to match the metal s, p, and d_σ orbitals:

1. Ψ(a_1g) = (1/√6)(σ₁ + σ₂ + σ₃ + σ₄ + σ₅ + σ₆) — overlaps head-on with the metal 4s orbital to form a strongly bonding 1a_1g MO and a high-energy antibonding 2a_1g* MO.
2. Ψ(t_1u) = { (1/√2)(σ₁ − σ₂), (1/√2)(σ₃ − σ₄), (1/√2)(σ₅ − σ₆) } — overlaps with the metal 4p_x, 4p_y, 4p_z orbitals to form three bonding 1t_1u MOs and three high-energy antibonding 2t_1u* MOs.
3. Ψ(e_g) = { (1/√12)(2σ₅ + 2σ₆ − σ₁ − σ₂ − σ₃ − σ₄), ½(σ₁ + σ₂ − σ₃ − σ₄) } — overlaps directly with the metal 3d_z² and 3d_x²−y² orbitals to form two bonding 1e_g MOs (predominantly ligand character, filled by ligand lone pairs) and two σ-antibonding 2e_gESCAST MOs (predominantly metal 3d_ character).

The Frontier d-Block Manifold: Non-Bonding t_2g vs. σ-Antibonding e_g*

Notice that Γ_σ contains no t_2g component. In a complex with pure σ-donor ligands (such as NH₃, which has no available p_π lone pairs or low-lying πESCAST orbitals), the three metal t_2g orbitals (dxy_, d_xz, d_yz) have zero net overlap with the ligand σ framework and remain strictly non-bonding at their unperturbed metal atomic orbital energy. Meanwhile, the twelve electrons donated by the six ligand lone pairs completely fill the six low-lying σ-bonding molecular orbitals (1a_1g² 1t_1u⁶ 1e_g⁴).

The metal's own dⁿ valence electrons therefore occupy the frontier manifold consisting of the non-bonding t_2g triplet and the σ-antibonding e_g* doublet. In Ligand Field Theory, the octahedral splitting energy Δ_o is simply the energy separation between the non-bonding (or π-modified) t_2g level and the σ-antibonding e_g* level:

Δ_o = E(e_g*) − E(t_2g)

Because e_g* is a covalent σ-antibonding orbital, placing an electron into d_z² or d_x²−y² directly weakens and lengthens the metal–ligand σ bonds along the corresponding axes. This explains immediately why uneven occupation of e_gESCAST (e__g¹ in high-spin d⁴ Cr²⁺/Mn³⁺ or e_g³ in d⁹ Cu²⁺) drives a massive 0.20–0.40 Å Jahn–Teller distortion (and why solid-state oxides undergo the cooperative Jahn–Teller effect), whereas uneven occupation of the non-bonding t_2g shell causes virtually no bond-length distortion.

π-Donor vs. π-Acceptor Ligands and the Origin of Δ_o

When ligands possess orbitals of π symmetry perpendicular to the M–L bond axis (twelve ligand π orbitals in total, transforming in O_h as Γ_π = t_1g + t_2g + t_1u + t_2u), the t_2g SALC interacts directly with the metal t_2g (d_xy, d_xz, d_yz) orbitals. Depending on whether the ligand π orbitals are filled low-energy lone pairs or empty high-energy π* antibonding orbitals, π-bonding shifts the metal t_2g level in opposite directions:

1. π-Donor Ligands (Weak-Field Halides, Oxo, Hydroxide, Sulfide): Ligands such as I⁻, Br⁻, Cl⁻, F⁻, OH⁻, and S²⁻ possess filled p_π lone pairs that lie lower in energy than the metal 3d(t_2g) orbitals. Mixing these filled ligand p_π(t_2g) orbitals with the metal d(t_2g) orbitals creates a lower-energy bonding π MO (filled by the ligand lone pairs) and pushes the metal-centered t_2g* frontier orbital upward in energy (making it π-antibonding). Raising E(t_2g*) narrows the energy gap to e_g*, decreasing Δ_o and stabilizing high-spin configurations.
2. σ-Only Ligands (Intermediate-Field Ammonia, Alkylamines, Hydride, Alkyls): Ligands such as NH₃, ethylenediamine (en), and CH₃⁻ lack p_π lone pairs or accessible π* orbitals. Here, t_2g remains purely non-bonding, and Δ_o is determined solely by the strength of σ-donation raising E(e_g*).
3. π-Acceptor Ligands (Strong-Field CO, CN⁻, NO⁺, Phosphines, Bipyridine): Unsaturated ligands such as carbon monoxide (CO), cyanide (CN⁻), nitrosyl (NO⁺), and 2,2′-bipyridine (bpy) possess empty, low-lying πESCAST antibonding orbitals (or empty 3d/σESCAST orbitals in PR₃) that lie slightly higher in energy than the metal 3d(t_2g) orbitals. Mixing the metal t_2g orbitals with the empty ligand πESCAST(t__2g) SALCs—a synergistic process called π-backbonding (or d_π → p_πESCAST back-donation)—pushes the metal-centered t__2g frontier orbital downward in energy (making it π-bonding) while shifting electron density from the metal onto the ligand. Lowering E(t_2g) dramatically widens the energy gap to e_g*, increasing Δ_o and enforcing low-spin configurations.

This three-part MO classification provides a rigorous first-principles explanation of Tsuchida's empirical spectrochemical series:

I⁻ < Br⁻ < S²⁻ < SCN⁻ < Cl⁻ < F⁻ < OH⁻ < H₂O < NCS⁻ < NH₃ < en < bpy < phen < NO₂⁻ < PR₃ < CN⁻ < CO

(π-donors: small Δ_o, high-spin)  →  (σ-only: medium Δ_o)  →  (π-acceptors: large Δ_o, low-spin)

Parameterizing Local σ and π Perturbations: e_σ and e_π

To apply Ligand Field Theory quantitatively to low-symmetry complexes—including trigonal bipyramidal, square pyramidal, distorted tetrahedral, and protein-bound active sites of arbitrary molecular geometry—Christian Klixbüll Jørgensen and Claus Erik Schäffer introduced the Angular Overlap Model (AOM) in 1965. Based on second-order perturbation theory, the Angular Overlap Model expresses the covalent destabilization (or stabilization) of each metal d orbital as a sum of independent pairwise interactions with each ligand L characterized by two radial parameters:

• e_σ (> 0): The energy by which a metal d_z² orbital is raised via head-on σ-antibonding overlap with a single ligand lying directly on the z-axis (E_σ ≈ H_ML,σ² / ΔE_ML).
• e_π: The energy shift of a metal d_xz or d_yz orbital interacting sideways (π-fashion) with a single ligand on the z-axis. For π-donor ligands, e_π > 0 (destabilizing/antibonding); for pure σ-donors, e_π = 0; and for π-acceptor ligands, e_π < 0 (stabilizing/bonding).

Summing the angular overlap factors over all six ligands of a regular octahedron ML₆ yields exact analytical expressions for the orbital energies:

E(e_gESCAST) = 3e__σ     and     E(t_2g) = 4e_π

Consequently, the octahedral ligand-field splitting parameter Δ_o is given directly by:

Δ_o = 3e_σ − 4e_π

For a tetrahedral complex ML₄ (T_d symmetry), the same AOM summation yields Δ_t = (4/9)(3e_σ − 4e_π) = (4/9)Δ_o, recovering the classic crystal-field geometric ratio while incorporating full σ and π covalency.

Synergistic Backbonding in Carbonyls and Heme O₂ Binding

The covalent framework of Ligand Field Theory is indispensable for understanding how metalloproteins bind diatomic gases (O₂, NO, CO) and why carbon monoxide is a potent physiological poison:

• Synergistic σ/π Bonding in Metal Carbonyls: When CO coordinates to a low-valent or electron-rich iron center, the carbon 5σ lone pair donates electron density into the empty metal d_z²(e_g*) orbital, while the filled metal d_xz and d_yz(t_2g) orbitals back-donate electron density into the empty CO 2πESCAST antibonding orbitals. Because populating the CO 2πESCAST orbital reduces the C–O bond order from 3 toward 2, stronger metal-to-ligand π-backbonding directly lowers the C–O infrared stretching frequency ν(CO) from 2143 cm⁻¹ in free gas-phase CO down to 1900–2050 cm⁻¹ in terminal metal carbonyls and carboxyhemoglobin (HbCO).
• Hemoglobin and Cytochrome P450: In deoxyhemoglobin, five-coordinate Fe²⁺ (d⁶) sits in a weak-to-intermediate ligand field in a high-spin (S = 2, t_2g⁴e_g*²) state where population of the σ-antibonding d_x²−y² orbital expands the Fe²⁺ ionic radius (0.78 Å), forcing the iron atom ~0.4 Å out of the porphyrin plane. When strong-field π-acceptor O₂ (or CO) binds at the sixth axial site, strong σ-donation and d_π → O₂(π*) backbonding increase Δ_o across the spin-crossover threshold, vacating the σ-antibonding e_gESCAST orbitals (t_2g⁶e_gESCAST⁰, diamagnetic S = 0), shrinking the iron radius to 0.61 Å, and snapping the Fe ion into the heme plane to trigger the T → R allosteric transition.

How does Ligand Field Theory differ from Crystal Field Theory?

Crystal Field Theory (CFT) is a purely electrostatic model that treats ligands as classical negative point charges (or point dipoles) that repel electrons in the metal d orbitals without any orbital overlap or electron sharing. While CFT correctly predicts the symmetry labels (t_2g and e_g) of the split d orbitals from group theory, it cannot explain why neutral π-acceptor ligands like CO and PPh₃ produce much larger Δ_o values than anionic halides, nor why interelectronic repulsion parameters (B) decrease in complexes compared to free gas-phase ions (the nephelauxetic effect). Ligand Field Theory (LFT) incorporates covalent molecular orbital overlap (σ-donation, π-donation, and π-backbonding), showing that e_g* is a σ-antibonding MO and t_2g is shifted by π-bonding (Δ_o = 3e_σ − 4e_π).

What is the 18-electron rule in Ligand Field Theory?

In organometallic complexes with strong-field σ-donor/π-acceptor ligands (such as Cr(CO)₆, Fe(CO)₅, or ferrocene Fe(η⁵-C₅H₅)₂), the octahedral splitting Δ_o is so large that the σ-antibonding e_gESCAST orbitals (plus the higher 2a_1gESCAST and 2t_1uESCAST orbitals) are pushed to prohibitively high energies. An octahedral complex achieves maximum thermodynamic stability when all six low-lying σ-bonding MOs (1a_1g² 1t_1u⁶ 1e_g⁴ = 12 electrons) and all three π-bonding t_2g⁶ MOs (6 electrons) are completely filled, while the antibonding e_gESCAST⁰ orbitals remain completely empty: 12 + 6 = 18 valence electrons.

What are Ligand-to-Metal and Metal-to-Ligand Charge Transfer (LMCT and MLCT) transitions?

Whereas weak d–d transitions (Laporte-forbidden, molar absorptivity ε ≈ 1–100 M⁻¹ cm⁻¹) involve exciting an electron between the metal-centered t_2g and e_g* frontier orbitals, charge-transfer transitions are Laporte-allowed electric-dipole transitions (ε ≈ 2,000–50,000 M⁻¹ cm⁻¹) between predominantly ligand MOs and predominantly metal MOs. Ligand-to-Metal Charge Transfer (LMCT) excites an electron from a filled ligand σ or π bonding orbital into an empty metal t_2g or e_g* orbital—producing the intense purple color of permanganate ([MnO₄]⁻, d⁰) and the vivid 600 nm blue band of plastocyanin (Cys84 S_π → Cu²⁺ d_x²−y²). Metal-to-Ligand Charge Transfer (MLCT) excites an electron from a filled metal t_2g orbital into an empty ligand π* orbital—producing the intense red-orange color and photoredox activity of [Ru(bpy)₃]²⁺.

1. Van Vleck, J. H. (1935). Valence strength and the magnetism of complex salts. The Journal of Chemical Physics, 3(12), 807–813. https://doi.org/10.1063/1.1749596
2. Griffith, J. S., & Orgel, L. E. (1957). Ligand-field theory. Quarterly Reviews, Chemical Society, 11(4), 381–393. https://doi.org/10.1039/QR9571100381
3. Schäffer, C. E., & Jørgensen, C. K. (1965). The angular overlap model, an attempt to revive the ligand field approaches. Molecular Physics, 9(5), 401–412. https://doi.org/10.1080/00268976500100551
4. Ballhausen, C. J. (1962). Introduction to Ligand Field Theory. McGraw-Hill. (Reviewed in Journal of the American Chemical Society, 85(8), 1210. https://doi.org/10.1021/ja00891a050)